Show HN: A shell CLI tool powered by LLM and RAG (include data security tips)

Re:Show HN: A shell CLI tool powered by LLM and RAG (include... by huangyz0918 on Tuesday April 16, 2024 @06:57PM (#40058337)

Thank you for sharing this!

Information security is an important issue for applying RAG! No one wants their personal information to go to OpenAI's server, but all users want to have a personalized assistant. -- That's the reason we need guardrails. A good framework you can take a look at: https://github.com/guardrails-ai/guardrails

But, it may not be able to integrate with small LLM agents easily.

[ Parent | on HN ]

Re:Show HN: A shell CLI tool powered by LLM and RAG (include... by cheeseisinsane on Tuesday April 16, 2024 @07:58PM (#40058849)

To make the best of LLM and RAG, the need for more data is required. Our next step will be focusing on enhancement of data protection and security through the integration of good frameworks such as guardrails and others in our system, while maintaining a promising performance in the task.

[ Parent | on HN ]

Re:Show HN: A shell CLI tool powered by LLM and RAG (include... by huangyz0918 on Tuesday April 16, 2024 @08:39PM (#40059136)

Maybe add local LLM support

[ Parent | on HN ]

Re:Show HN: A shell CLI tool powered by LLM and RAG (include... by cheeseisinsane on Tuesday April 16, 2024 @09:18PM (#40059359)

of course. will add that in a short future

[ Parent | on HN ]